Sparse Signal Processing Technology Research For Large-Scale Wireless Access

Random access model and user identification algorithm of Long-term Evolution can not meet large-scale random access of Machine Type Communication.,because its only complete user identification in random access process and limit the number of user in system.So in this paper,in order to finish the user identification,at the same time complete the channel estimation in the random access,we design a new type of random access model and algorithm based on compressed sensing and characteristics of large scale service of MTC.First of all,this paper introduces the limited scope of LTE random access and put forward that we need to solve the problem as follows,high efficiency,The design of new type of random access model and the design algorithm that allows the system to accommodate more users and allows complete user identification and channel estimation at the same time.Secondly according to the characteristics of MTC service,taking finishing the user identification,completing the channel estimation at the same time as the main purpose,solving the problem of random access by means of compressed sensing technology.We analyze the necessity of double sparse for user identification and channel estimation aiming at the synchronization,frequency selective fading channel.based on the intensive active user sharing the common sparse structure,the effective GOMP-UT algorithm is put forward.For synchronization,double selective fading channel,based on different OFDM symbols sharing the common sparse structure,we design model and put forward effective Union-OMP algorithm that reduce the algorithm complexity.Based on asynchronous,line-of-sight channel and rayleigh channel,when the active user asynchronous delay is unknown,we design model that can make us to complete user identification and channel estimation,and design the corresponding algorithm,also put forward the AOMP algorithm based on the line-of-sight channel.Then we use multiple response sparse bayesian algorithm by the analysis of structure sparse characteristic of model to complete the user identification and finish the channel estimation at the same time for SCMA uplink.Also,we analyze the control signal generated in our model.The M-SBL has low complexity by complexity analysis,and has effectiveness by simulation result and analysis.Finally,aiming at synchronous frequency selective fading channel,with sparse Bayesian algorithm and CDMA multiplexing method,we establish the random access model using USRP.We assume four active users and one central node.The link established using USRP is for testing the model and the algorithm that using compressed sensing to solve the problem in random access procedure is the practical feasibility.The next generation mobile communication system needs accommodate more users,and needs ensure high efficiency and low ductility on the basis of high reliability.So in view of the MTC service,it is necessary to design new type random access model and corresponding algorithm.